Jump to content


Photo

USB frame grabbers: Is there a better solution?

  • Please log in to reply
5 replies to this topic

#1 scopenitout

scopenitout

    Vostok 1

  • -----
  • Posts: 101
  • Joined: 24 Aug 2013
  • Loc: Mt. Belzoni

Posted 24 January 2014 - 12:16 PM

I'm using a Mallincam Xtreme with the Mallincam MCV-1 USB frame grabber. I also have a Dazzle USB grabber. I don't find much difference in image quality between the two devices.

Why is the image quality straight from the camera to a dedicated monitor, so much better than the image produced by the grabber sent to my laptop?

I like using Miloslick software to enhance the image, but it cannot do much of it on the raw, direct from camera signal.

What is it with these USB grabbers that so degrade the image?
Is there a hardware solution that will give better image quality with full, software image enhancement functionality?

#2 Peter D.

Peter D.

    Mariner 2

  • *****
  • Posts: 216
  • Joined: 09 Feb 2012
  • Loc: Central New York

Posted 24 January 2014 - 01:08 PM

There's random noise in the video signal that's different in every field. The monitor smooths it out a little bit because its phosphors have "latency" (retaining some information from the previous few frames), and your eye does the rest of the integration at the 1/60 second rate of the video display (your eye can't see changes at that rate, so everything is averaged). That is, the combination of the monitor and your eye is stacking the image, accomplishing noise reduction and sharpness enhancement simultaneously.

But the frame grabber samples just one frame, and it adds noise of its own (both random and quantization noise) on top of the noise that's already there.

Really good on-the-fly post processing could accomplish the same thing, but the state-of-art isn't so good that substantial additional delay is incurred, and significant extra skill is required.

Although it may be possible to build a "perfect" frame grabber (at substantial additional expense), that can't help the fact that only one frame is converted at a time: post-processing is required for a complete solution.

Although software is getting better at on-the-fly post processing, it still can't match what the combination of a monitor and your eye/brain can accomplish.

Of course more sophisticated post processing is capable of achieving much more than any videocam/monitor/eye can do (taking substantial time, requiring great skill, and adding stress and frustration), but that's "imaging", which does not fall under the scope or intention of this forum.

I like it that way: although I also image when I have the time and patience, instant gratification is the greatest! Anything more than 10 seconds is torture, and 1 second is great!!!

#3 scopenitout

scopenitout

    Vostok 1

  • -----
  • Posts: 101
  • Joined: 24 Aug 2013
  • Loc: Mt. Belzoni

Posted 24 January 2014 - 04:19 PM

Peter...
I can't thank you enough for your thorough explanation.
As a relative newcomer to Video Assisted Observing, I have a lot to learn. As far as "instant gratification" (of Video)goes, I'm with you all the way! I haven't opened my eyepiece case since I went to Video.

#4 Relativist

Relativist

    Skylab

  • *****
  • Posts: 4017
  • Joined: 11 Oct 2003
  • Loc: OC, CA, USA

Posted 25 January 2014 - 12:04 AM

I'm of the same opinion as yourself, I sat on the sidelines not wanting to jump in because I wanted to completely avoid the frame grabber. I've given up on that, for now.

#5 Peter D.

Peter D.

    Mariner 2

  • *****
  • Posts: 216
  • Joined: 09 Feb 2012
  • Loc: Central New York

Posted 25 January 2014 - 02:36 PM

Unfortunately a video monitor is not an option for me since I'm trying to set up a remote observatory, where all viewing must be done through the computer interface, over the internet. Even if I could digitize and transmit the video in real time, the limited bandwidth available over a conventional internet connection (uplink speed is severely limited) couldn't handle it without complex MPEG compression. So I'm not giving up on the frame grabber.

For a quick look I accept the limitations of a quick single-frame grab, but I'm currently researching options for on-the-fly processing that automatically averages and/or stacks multiple frames in real time for the virtual "live view" that I seek.

#6 Raginar

Raginar

    Fly Me to the Moon

  • *****
  • Posts: 6138
  • Joined: 19 Oct 2010
  • Loc: Rapid CIty, SD

Posted 25 January 2014 - 09:18 PM

Peter,

You can do what you're trying to do with a frame grabber over VNC. You're not losing anything and VNC doesn't require a crazy capable internet connection.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics